I've been reading Francis Hamit's book The Shenandoah Spy, and I'm about half way through it. I found out about this book via Hamit's postings at other blogs, where he mentioned that it had been shopped to commercial publishers with little success. I don't claim to be an expert in the publishing industry, but I think I know why the book didn't get picked up.
The first chapter is slow. Now, please don't get me wrong - the book is both interesting and well-written, and I'm glad I bought it. But, Hamit started his story at the beginning of the war, and, well, not much happens the first chapter or so.
The book I read previous to Shenandoah Spy was Tanya Huff's military SF novel Heart of Valor. In this book, the central issue, a mystery, opened up in chapter 1, and things get off to a ripping fast start.
This is a new-ish phenomenon for commercial fiction. I've read books written as late as the 1970s, books billed as "action-packed," and they started more deliberately. Obviously, exceptions abound in either direction, but my point is this. Commercial publishers are looking for stuff they think will sell. One of the key factors they use to make that decision is how quickly one is pulled into the story.
Part of what got me thinking about this is that, as I was reading Shenandoah Spy, a scene came up where my inner editor said "this would be a great first chapter." But to use the scene as a first chapter would require telling part of the story as a flashback. Hamit apparently decided he didn't want to do that.
Like I said, Shenandoah Spy has so far been an enjoyable read. But as a writer, the book points out one of the advantages of self-publishing: the ability to ignore conventions of commercial fiction. This can lead to interesting things.